Un informe europeo revela que la formación en inteligencia artificial infringe los derechos de autor
A joint study by technology and legal experts concludes that training generative artificial intelligence constitutes a direct infringement of copyright. By using unfiltered data from the internet, these models do not benefit from legal exceptions such as text mining; instead, they memorize and can reproduce protected content, amounting to a massive theft of intellectual property. This finding challenges the current technological narrative, demonstrating that innovation should not be built on the illegal exploitation of human creative effort. This revelation is crucial for the open data movement, as it exposes the fundamental contradiction between the culture of free access to information and intellectual creation rights. This case illustrates how the massive availability of online data does not automatically imply its legality for commercial or algorithmic use. By making the “black box” of data processing visible, the report forces a reconsideration of how open data is stored, processed, and shared, emphasizing that technical transparency must be accompanied by strict respect for intellectual property laws. Finally, this finding underscores the need for clearer legal frameworks that balance technological innovation with the protection of creators. The ball is now in the legislators’ court to establish ethical and legal boundaries for data use. For the open data community, this means that the sustainability of open data initiatives depends on integrating mechanisms that safeguard copyright, ensuring that access to information does not become a vehicle for the misappropriation of human knowledge.
Source: lavozdegalicia.esPublished on 2024-09-22